Traditional APS replacement
This is the third and final paper in our series presenting the findings from Munich Re’s studies on the protective and operational value of electronic health records (EHRs) in the underwriting process.
Our research is based on the premise that EHRs have improved the amount and quality of information they capture and EHR use is becoming more common as a growing number of carriers integrate them into their underwriting processes.
To understand these enhancements and how they can be utilized, we completed a series of large-scale, wide-ranging retrospective studies to find the protective and operational value of EHRs in multiple underwriting scenarios. The study included a sample of over 800 underwriting files consisting of business from multiple carriers that target middle to high-net-worth markets.1 Broker, general agency, and direct-to-consumer distribution channels were included, as were traditional and fluidless underwriting paths.
The use cases studied included:
Adding EHRs to accelerated underwritten cases, i.e., no labs or attending physician statement (APS)
The primary findings in this use case highlighted the effectiveness of EHRs as a core underwriting tool, demonstrating their value in accelerated underwriting for both point-of-sale and post-issue audit use cases.
Using EHRs to replace fluids for fully underwritten business without an attending physician statement (APS)
This study also revealed the effectiveness of EHRs as a core underwriting tool, showing their potential value in fully underwritten non-impaired use cases without an attending physician’s statement.
Assessing the impact of replacing traditional APSs with EHRs (this paper)
Background and data
Our third study focuses on 303 cases of the 800 that had both APSs and EHRs available. Cases that were missing either an EHR or APS were excluded. Face amounts were limited to greater than or equal to $100,000, and issue ages were limited to greater than or equal to 18. These face amount and issue age filters brought the study case count down to 301. The sample included middle and high-net-worth markets with distribution representing brokers, general agents, and direct-to-consumer channels.
To conduct this study, an underwriter reviewed cases using disclosures (applicant admitted history), MIB check, prescription (Rx) and/or medical claims (Dx), insurance exam + labs, and an EHR (EHR scenario). Then they reviewed the same case with the addition of an APS (EHR + APS scenario). To ensure an independent review, we engaged external and internal underwriters to help complete the risk assessments. For consistency, they utilized the Munich Re EDGE underwriting manual to assess each risk in the scenarios indicated.
Methodology
- Table 1 shows the relative mortality assumptions by risk class that were used in the analysis.
A count-based average expected mortality was calculated for both the EHR scenario and the EHR + APS scenario, based on the relative mortality factors outlined above. There were policies for which an assessment could not be made given the available underwriting evidence in the EHR scenario and/or the EHR + APS scenario. These undetermined underwriting risks were excluded from the mortality calculation, and therefore, the average expected mortality includes all policies that received an assessment in that scenario.
Assessments could have been considered undetermined for any reason the underwriter did not believe there was enough data to make a decision, such as cases with incomplete or old data. During the application process, an undetermined assessment would usually be followed by a request for more information, medical or otherwise.
The mortality saving was calculated as follows:
Mortality saving % = Average expected mortality for EHR / Average expected mortality for EHR + APS – 1The mortality saving was further quantified in dollars per policy by projecting a present value of future death benefits (PVDB), which was used as a surrogate for the present value of premiums. The 2015 Society of Actuaries (SOA) Valuation Basic Table (VBT) with no mortality improvement was used as the underlying mortality basis. Lapse and interest rates were set to best estimate assumptions. Projection length was set to a 10-year time horizon.
Results
The study results and analysis are viewed through the lens of an EHR-first strategy. This means underwriting these cases with an EHR first and then layering on an APS to determine if the risk assessment changed.
Of the 301 cases eligible for inclusion in the study:
- In 221 cases (73%), the EHR provided sufficient information for a decision to be made (Table 2).
- In 68 cases (23%), the EHR did not provide sufficient information, and the addition of the APS enabled an underwriter to make a decision.
- In 12 cases (4%), even after the addition of the APS, an underwriter was unable to make a decision.
The aggregate mortality savings are -7%. The negative number implies that the EHR-first strategy introduces mortality slippage, which is the opposite conclusion of our previous two studies.
Two additional takeaways to note are the fact that in 73% of cases, EHRs contain sufficient information for a decision to be made. Then, given an EHR contains sufficient information, the decision is unchanged with the addition of an APS 87% of the time. This indicates that for a high percentage of cases, the EHR contains all the necessary information to make a decision, and given we are decisioning with an EHR, the decisions we make are in agreement with those from APSs the large majority of the time.
Analysis
Mortality implications
On the 221 cases that we could decision on using EHRs, we arrived at the same decision as the EHR + APS scenario in 87% of the cases. Looking at the risk classes of these 87%, we see over half are Standard. The Standard cases mostly consisted of an APS for age/amount or ordering an APS to evaluate an MIB code, Rx/Dx hit, exam disclosures (e.g., mental nervous, obstructive sleep apnea), and lab findings.
This identifies a subset of applicants who would have previously received an APS where the EHR first approach performs just as effectively.
In reference to the 46 substandard risk cases above, there were a host of reasons for ordering APSs, including application disclosure information, exam/lab results, MIB codes, and Rx findings. From an impairment standpoint, the leading ones are listed in Table 4.
The majority (57%) of the differences in decisions between the two scenarios are a matter of rating magnitude. These 57% of cases could be issued in both scenarios. Nine cases show a greater discrepancy and would not be issued in the EHR + APS scenario; they consist of four postpones and five declines. In only one case did the EHR not identify tobacco usage where the APS did.
Table 6 summarizes the most prevalent impairments for the 23 cases where EHR + APS had a higher decision rate than EHR. The most frequent primary impairment was cardiovascular, followed by build/BMI and diabetes. There were a few cases where the EHR did not provide information on the primary impairment. These cases represent a subset of cases where a more judicious ordering of APSs could provide material slippage protection. In our study, we encouraged underwriters to make a decision with the information provided. We see a handful of cases where a production underwriter could have opted not to decision with the information as is because the EHR was lacking the specific information they were looking for.
Of the cases where we could not make a decision using an EHR but could with EHR + APS, 45% are either postpone or decline cases. This is indicative of the fact that the policies in this pool that could not be decisioned with EHRs are notably more impaired lives and complex cases.
This is also illustrated in Table 8, which looks at the most prevalent impairments in the 68 cases. We see that mental nervous is the most common impairment, which isn’t surprising, as psychological assessments and evidence often require a higher level of authorization and aren’t as available in EHRs as they are in APSs.
Underwriter preference
Evidence cost
Aside from mortality implications, there are evidence-cost implications associated with an EHR-first underwriting approach. For all of these 301 cases, participating carriers determined that acquiring an APS was required to underwrite the case. Due to data limitations we can’t say for certain what percent of these 301 cases had an APS ordered for cause as opposed to age and amount requirements. We do know that the participating carriers chose to order an APS for all 301 of these cases for one reason or another.
If we assume that the average cost of an APS is $80, the baseline evidence cost would have been: 301 * $80 = $24,080.2
In the EHR-first approach, 221 of the cases required only an EHR, and 80 would have had an EHR and an APS: 221 * $55 + 80 * ($55 + $80) = $22,955.3
Immediately, we can see that the EHR-first approach breaks even on an evidence-cost perspective by $1,125, or about $4 per case.
This analysis does not take into account the time savings from ordering EHRs and the associated knock-on effects. Approximately 50% of EHRs are typically delivered within one hour, 80% within 24 hours, and 90% within 48 hours. APSs can frequently take weeks. This has strong implications for the customer experience and can decrease the number of applicants who choose not to continue their application due to hold-ups in the underwriting process. These cases represent lost hard dollars in terms of evidence costs, distribution agents’ sunk time, and lost future revenue and profit from policies not issued.
Conclusions
The results of this comprehensive EHR study reveal several opportunities in the underwriting process where EHRs can be an effective tool, specifically, using EHRs in lieu of APSs for age/amount requirement and as a follow-up tool on MIB and Rx/Dx inquiries. In addition, EHRs can be an excellent tool for many single and dual medical histories and exam disclosures/findings, including build/BMI, type 2 diabetes, lab abnormalities, obstructive sleep apnea, mild mental nervous conditions, and mild cardiovascular impairments, to name a few. It is also worth mentioning that EHRs provided more up-to-date information than APSs in several cases.
The fact that we see negative mortality savings in the EHR-first strategy doesn’t come as a surprise to us. We aren’t advocating for an outright replacement of APSs with EHRs. Each carrier will balance mortality slippage, evidence cost, and customer experience differently. Slippage can be controlled through careful underwriting and a judicious approach to ordering an APS when presented with an EHR that might be relatively thin on data or not contain medical data from a particular provider.
There are 221 cases that could be decisioned using EHRs, and in 192 (87%) of those cases, the addition of an APS did not change the decision. However, our analysis also shows that for the majority of these 192 cases, underwriters preferred the more detailed information provided by an APS over an EHR. This preference is understandable, given underwriters' familiarity with APSs and their desire for comprehensive information. Speaking with our study underwriters, we also heard that some simply prefer APSs. For others, it didn’t matter. Adapting to alternative sources of information, such as EHRs, requires a mindset adjustment. This shift is similar to underwriters learning to rely on APS summaries instead of full APSs, or underwriting managers trusting automated underwriting (AUW) requirements in place of some traditional ones. Ultimately, this adjustment must be balanced with the need for more detailed information in higher-risk cases, ensuring that underwriters have the necessary data to make informed decisions while also embracing the efficiencies offered by alternative information sources.
References
Contact the authors
Related Solutions
Newsletter
properties.trackTitle
properties.trackSubtitle